Ionospheric products from sensors and models were compared to investigate strengths and limitations of each. Total electron content data from computerized ionospheric tomography (CIT) and TOPEX sensors in the Caribbean region in 1997 were compared to estimates produced by models Parameterized Ionospheric Model (PIM) and Raytrace/ICED-Bent-Gallagher (RIBG) and global maps from GPS. A 5 total electron content unit (TECU) bias was observed in TOPEX. CIT and TOPEX confirmed the location and structure of the equatorial anomaly. A GPS map confirmed the location of the anomaly but did not reproduce structure less than 1000 km in latitude and 1500 km in longitude and underestimated TEC by at least 11 TECU or 25%. PIM positioned the anomaly 13¿ equatorward of its observed location and greatly underestimated (~50%) the rise in content over 5¿--25 ¿N range. RIBG overestimated the latitudinal extent of the anomaly and underestimated TEC at the peak by 40%. Additional comparisons were made using CIT and ionosonde sensors at midlatitude during the summer of 1998. Fourteen days of TEC, hmF2, NmF2, and half-thickness comparisons showed reasonable agreement between CIT and ionosonde for TEC and NmF2. The hmF2 and half-thickness comparisons were contaminated by noise, which accounted for a significant portion of the ionospheric variation. Daytime cases where CIT overestimated maximum density were attributed to underestimating layer thickness. Finally, TOPEX and multiple GPS sensors were compared to verify regional ionospheric conditions associated with occurrence of nighttime ionospheric depletions in the Caribbean during Combined Ionospheric Campaigns in June of 1998. From 0300 to 0800 UT on June 26, GPS and TOPEX showed elevated nighttime content over the entire Caribbean region. Vertical TEC approached 25 TECU in some places with interspersed depletions, which in some cases evacuated nearly the entire ionospheric content. ¿ 2001 American Geophysical Union