Getting rid of data outliers

[mwwoodm]

I have some mostly good data with some extreme outliers on the high side. My goal is to find the maximum value of the good data. My idea is to use a function similar to the MEDIAN function that will pick out the 90th percentile member of the population. MEDIAN does a great job for the 50th percentile member and PERCENTILE appears to give a percentile fractional number of the inferred distribution. Any way that I could get these two functions to get together and make a new function that picks the highest non-outlier member from my data?

Thanks in advance!