Scanning electron microscopic studies of silica valve formation in naviculoid diatoms representing six different genera revealed that the precise sequence of depositional events varied among genera. Valve deposition begins with the formation of the raphe sternum, from which virgae (lateral outgrowths) extend. Areolae (pores) are formed between the virgae by the fusion of cross-extensions (vimines). In most of the species studied (Craticula ambigua (Kützing) D. G. Mann, Frustulia vulgaris (Thwaites) De Toni, Craspedostauros australis E. J. Cox, and Gomphonema truncatum Ehrenberg), areola (pore) formation began near the raphe sternum before completion of the valve margin, but in Pinnularia gibba Ehrenberg the valve margin fused before the areolae were formed. Silica deposition in all these taxa was mainly distal to proximal (with respect to the cytoplasm), but in Haslea sp. it was mainly proximal to distal. Haslea also differed in that areolae were defined as the valve margin was completed. These data have also contributed to the interpretation of taxonomically important features, such as raphe endings. In P. gibba the internal central raphe fissures were laterally deflected but subsequently obscured by additional silicification of the valve, whereas in G. truncatum they were initially straight, becoming laterally deflected as valves mature. External raphe fissures in Frustulia became Y-shaped only just before maturity; in immature valves they were dotlike, as in Amphipleura Kützing. The comparison of developmental pathways in diatoms is a useful adjunct to morphological and other approaches in diatom systematics and warrants renewed attention.