The continuous series of sea level at Key West, Florida commenced in 1913, but we have discovered sporadic measurements that date back to 1846. From records at the U.S. Army Corps of Engineers and the U.S. Coast and Geodetic Survey, the sea level series has been connected to a Summary (common) Datum. Thus, a gappy record of monthly and annual mean heights (H[t]), perhaps the United States' longest series over San Francisco (ca. 1854) or New York ( ca. 1856), can be tested to ascertain if the rise in relative sea level at this site is stationary. Applying first and second order least squares and two-phase regression analyses, we find that dH/dt is 0.19±0.01 cm/yr, and that d²H/dt²= [9.6±8.6]·10−3 cm/yr²; the two-phase regression shows H[t] rising 0.15±0.03 cm/yr before ca. 1925 and 0.23±0.01 cm/yr afterwards. Neither the second-order regression coefficient nor d²H/dt² nor the two-phase calculation are significant above the 75% confidence level, but all three are weakly consistent with accelerated rise. For the epoch 1951–1987, Key West sea level, corrected for post-glacial rebound, is best explained by concurrent measurements of 0–1,000 db dynamic height anomaly change.