Here are some tips from a former Science Fair judge in both the school and Regional Science Fair divisions.

Start early.  Start VERY early.  You know that you are going to have to do a project, so start thinking about it before school even starts in the Fall. 

Pick a project that interests you.  If you choose something that you are curious about or if you think that with study, you can find the answer to something that you have a question about, your project won't be so much of a chore. 
If you are curious about how pollution of coastal waters affects oyster beds, and you REALLY want to know, don't pick a project involving which oil additive really makes automobile engines run best.  When you are bored with a project, you won't do as well as someone who chose a topic that interested them.

In order to have a really good project that attracts judges' attention in a positive way, try to prove or disprove a hypothesis [(hi - poth' - uh - sis) --a supposition or theory that may be either proven or disproved through experimentation] rather than merely copying an experiment that has been done to death, such as the construction of a "hair hygrometer."   After the judges have seen TWO hair hygrometers, there has already been two too many...even if one uses unusual hair samples such as boar hair, elephant hair, horse hair...it is still a hair hygrometer.  All of the underlying principles are well-documented, and you haven't really added anything of significant value. 
If, however, you wanted to find out if there is a correlation with the moon's (Luna's) phases and unusual personal behavior (lunacy, crime, childbirth), you might be able to gather data in a method that demonstrates your hypothesis (either that Lunar phases do or don't affect human behavior) in a way that gives us a greater understanding, and which actually answers your question.

Some projects are set up to begin a study which is continued through successive science fair projects.  Before you try that kind of project, get some expert advice on the merits of your initial hypothesis, and whether or not it can hold up to continued research in the coming years.

Keep good and accurate records of your project.   If you are supposed to check an experiment and you miss the checkpoint time, log it as a miss...don't try to fudge the results on the report.  If your missing the checkpoint might alter the eventual results of the experiment, make a notation of what differences there might be in the results since you missed a critical moment; or, if there are no differences, list that, too.

Keep good and accurate records of any reference materials that you use...newspapers, magazines, books, personal interviews...whatever.

If necessary, purchase separate notebooks or folders in which to log your results.  Make them as organized and as neat as possible.  Do not decorate them with a lot of drawings, unless you are using graphical illustrations of various aspects of your project. 
If you begin your log a couple of pages from the front of a composition book, you will be able to add a Table of Contents and page numbers on the opening pages.  That makes it easier for judges to find specific things that they want to check, and is very impressive from a judging standpoint.

Use SPELLCHECKER.  Make sure that your grammar is correct.  Judges don't want to see entries such as, "When the jar broke, I seed that the hole thing had come a part, and my experminent was all runed."   Failure to prove your hypothesis as correct is not considered a failure to judges...especially if you can see where you went wrong and have a good idea of what to do to make it right in a future round of experimentation.  Judges do see failure to proof your work and correct mistakes as sloppy and unprofessional.  Poor spelling, which could be avoided by the use of a dictionary or Spellchecker, is not excusable.   Poor grammar can be checked, also.  Get someone to proof your work, if you need help in that area.   

Be honest.  If your hypothesis is that indoor plants cannot flower under fluorescent lighting, but must have access to either real sunlight or "Grow Light" complete spectrum sticks or light bulbs, and your test plant under the fluorescent lights flowers anyway, state in the conclusion that your hypothesis was wrong.   Being wrong doesn't make your project a failure.  It tells judges that you were thorough and that your work has disproved your own pet theory.  It happens in science all the time.  Even when you are wrong, you learn something.  Honesty impresses judges. 

Do not schedule the monitoring of an experiment for times when you will not be available to do the check yourself.  If a judge sees that you checked on an experiment at 6 AM, Noon, 6 PM, and Midnight every day, the obvious conclusion is that Mommy is the only one who should get credit for your project.  The judge knows that unless you were home schooled, you were at school part of the time, and probably in bed long before Midnight.  A few innocent questions regarding your extracurricular activities will alert the judges further when they find out that you take gymnastics/debate/band or whatever from 5 PM to 6 PM every Monday, Wednesday and Friday.   If you aren't available to monitor your project at the times necessary, choose a different project that YOU can do yourself.

Your Thesis Paper should include your working hypothesis, and an outline of what you plan to do to prove or disprove your hypothesis.  

Your Research Paper should include:  1) your working hypothesis (your thesis), 2) the body of detailed research and experimentation along with pertinent dates or dates/times, 3) your conclusion detailing whether or not you proved or disproved your hypothesis...and how you could improve your methodology or alter your hypothesis to one that is more easily proven, and 4) references (bibliography).   

If you get outside help such as a mentor from a high school, college, or university, give them credit as your mentor.  Be willing to discuss how they helped you.  If they, however, produced most of the actual results of your project, the judges will know.  Get guidance from experts if you need it, but do the work yourself.
Judges question science fair participants about their project.  If you have had too much help from other sources, it may become obvious when you can't answer technical questions about your research and experimentation.  Remember that the judges that will be evaluating your project are experts in the particular field in which your project is classified.

FOLLOW THE RULES.  If there are to be no live animal displays, soil samples, biohazardous materials, glass containers, or bones or other animal products, don't choose a project that demands such items to be displayed for judging.  Judges won't blame the Rules Committee for not allowing your complete project display to enter the fair...they will blame you for violating the rules to begin with.  Before your project is allowed into a Regional Science Fair, it has to pass inspection by people who know the rules.  If your school allowed you to violate the rules for the school fair division, your project, nonetheless, will be stopped at the door of the Regional Fair if it made it that far.  Not even parental tantrums and complaints of, "That's just not FAIR!" will save it from disqualification.   Each school receives a copy of the numerous rules.  Schools may choose to disregard the rules, but the regional project inspectors won't.  Some of them are, or have been, judges.  If you don't know the rules and your science teacher hasn't given you a copy of the rules, ASK for a copy.   

Flashy backboards, lots of color and fantastic fonts may be eye-catching, but that's not what the judges will consider as being relevant.   What they will be looking for will be:  1) a good opening hypothesis, 2) thorough research with bibliographical notation (author, book or magazine title, article title for magazines, date of publication, page number), 3) concise (clear and to the point) writing, a clear conclusion, and neatness (of displayed notebooks and papers, backboard, and yourself).  Typing or computer printouts on standard printer paper (as opposed to scroll-feed paper) is best.  If you do not have access to such equipment, you may want to check with your local junior college or one of the high school computer instructors, and get them to help you as a mentor.  In that way, you will have access to the tools that you need.

If you have photographs on your display board, make sure that they are staged with adequate lighting.  Use close-ups when necessary.   You may need to get outside help in getting good quality photos that are not washed out with a bright flash, or too dark to interpret.  If you need help, get it.   This is another reason that you start EARLY.  You have more time to get help where you need it.    

For project judging, it is not a good idea for a participant to show the judges how much contempt they might feel toward the older generation, or how they like to disregard cultural norms by wearing skimpy clothing, or baggy and ill-fitting clothing, or unusual make-up/hairstyles.  Judges, after all, are human.  They might want to judge a project solely on its merits, regardless of how arrogant or sullen the participant may show themselves.  However, if your appearance and/or attitude is too much off the norm...if you look sloppy or bored or outright sullen...the judges may automatically assume that your work is either sloppy...or someone else's.  You may have more questions to answer before the judges are convinced that you actually did all of the work yourself. 
This is not a good time to make an outlandish fashion statement.  Don't get bent out of shape over your "personal rights" or your "civil rights."  The BEST way to present yourself for judging is to appear to be neat.  You may not want to "conform," to a class standard.  So it isn't you...you can do it for that short amount of time.  People are people, and all things being equal, nothing is.  Judges, being human, come with their own preconceived attitudes and prejudices.  Everyone has prejudices...the judge might not approve of nose rings...the participant may not approve of anyone who doesn't approve of nose rings.....    You can't get around it, different people have different standards, so, for the time being, do whatever will make you look your best to the judges.   This is akin to a job interview.  You put your best foot forward...and not in boots with spikes down the sides. 
To be honest, it is best for you to not wear any jewelry, or anything else that might be distracting.  If the judges are distracted by your appearance, they will not have their full attention on your project.  That same distraction will hurt their eventual evaluation of your project when they begin comparisons.  If all they remember about you is your personal appearance, the project may as well have stayed home. 
Answering with, "Yes (or No), Sir," or "Yes (or No), Ma'am," is never out of place.  Politeness, respect, and courtesy go a long way with judges.   They might not remember many of the other participants, but if you are particularly forthcoming, polite, and respectful, you will win points with the judges.  Your project may not win, but your reputation will.  When it comes right down to it, the only thing that you have of real worth in life is your reputation.  Polish it, and make it shine.  

One of your major goals should be to become so familiar with your project that you can quickly and thoroughly answer any questions that the judges propose.  Do so in a friendly manner, and speak clearly.  Look judges in the eye when speaking to them.  Eye contact signals the fact that you are confident, and that you take full responsibility for your work. 
If you constantly look away from the judges or mumble, it is often interpreted as a possible sign of insecurity (so the judge wonders Why the participant is insecure...what part of the experimentation was not done, or done correctly), or worse...the possibility of dishonesty. 

Don't be afraid to say, "I don't know," if you don't know an answer.  Tons of blather and "beating around the bush" to disguise ignorance does not impress judges.   A simple, "I don't know," does impress judges...unless it is in answer to a critical part of your project that you should be familiar with....

Practice answering questions about your project with a friend.  Ask them questions about their project.  It sounds rather silly, but if you practice looking someone in the eye and answering questions for which you haven't prepared answers, you will be more confident during the judging interview.

After every single participant has been reviewed, judges compare notes with each other and vote on the projects that they felt were both the best projects and the best presented.  They may elect to interview their chosen top subjects again.  If you are interviewed more than once, turn on the charm.   You've been standing at your project for what seems like forever and you're tired, but muster your strength and be friendly and polite and eager to answer questions.   Your charm, combined with the worthiness of your project, may win out over a competitor's project.

Relax.  Judging interviews aren't the end of the world as we know it.  Remember that judges are just people.  If you have an innovative project that hasn't been done to death...or a real twist to an old problem, and you know your material (and, hopefully, something about related areas to your own project subject), and you appear neat, courteous and friendly, you and your project will stand out as winners, even if you are bested by another project/participant.  If you made some mistakes this year, find out what you can do to make corrections--there is always next year.....