Proposal success rates are calculated for HST Cycles 11 through 21 as a function of the gender of the Principal Investigator (PI). In each cycle, proposals with male PIs have a higher success rate, with the disparity greatest for Cycles 12 and 18. The offsets are small enough that they might be ascribed to chance for any single cycle, but the consistent pattern suggests the presence of a systematic effect. Closer inspection of results from Cycles 19, 20 and 21 shows that the systematic difference does not appear to depend on the geographic origin of the proposal nor does it depend on the gender distribution on the review panels. Segregating proposals by the seniority of the PI, the success rates by gender for more recent graduates (Ph.d. since 2000) are more closely comparable. There is also a correlation between success by gender and the average seniority of the review panel for Cycles 19 and 20, but not Cycle 21. We discuss these results and some consequent changes to the proposal format and additions to the HST TAC orientation process.