During FGGE 1979, from March to July, an extensive oceanographic experiment with ships and moored stations was carried out in the Somali Current. The development of the monsoon winds off Somalia during the time of that experiment is described in a comparative analysis of standard ship wind observations, moored buoy wind measurements, low-level cloud winds, and winds from land stations. The onset 1979 is found to be of the multiple type, with northward winds off Somalia beginning around May 5 but dying down into early June; the real onset of sustained high winds starts around June 10. Cloud level wind observation numbers off Somalia decrease drastically with the monsoon onset because of lack of clouds over the quickly developing cold upwelling areas. An intercomparison of cloud level and ship winds for the period May 16 to July 6 at five offshore points shows good agreement in directions but reduction of ship wind speeds against cloud level winds off northern Somalia after the onset, which may explained by the increased vertical wind shear due to high air stability over the upwelled water and by geostrophic shear due to the strong gradients of sea surface temperature. A comparison of 3-day averages of buoy winds measured at 3-m height 30 km offshore, but still inland from the ship lane, with ship winds for the period March 3 to June 10 showed good agreement in directions but lower buoy wind speeds, which could partly be due to sensor height difference and partly due to horizontal wind shear towards the coast. Coastal stations and wind buoys near the coast are found not to be good indicators of the monsoon onset further out in the open ocean.