This paper reports the results of two experiments. In the first, voice-key naming times were collected and in the second, keypress naming times were collected for 250 of the Snodgrass and Vanderwart (1980) pictures. The resulting naming times and correct naming rates were well predicted in multiple regression analyses by one or another measure of codability (name or concept agreement) and by age-of-acquisition ratings collected specifically for this study. Voice-key responses appeared to be somewhat more sensitive indicators of naming difficulty, although keypress responses did remarkably well. The Appendix presents the age-of-acquisition ratings, the trimmed vocal and keypress naming times, and the correct naming rates from the two experiments for the 250 pictures.The cognitive operations underlying picture naming have long been of both theoretical and practical interest. Cattell (1886) first noted the fact that reading words was faster than naming pictures, and speculated on the processes that made word reading faster than picture naming. The reading advantage has generally been attributed to one of two processes: (I) the highly overlearned association between stimulus and response produced by much practice, and (2) the fact that in many languages, pronunciation of the word can be accomplished via a fast, grapheme-tophoneme translation process. Theios and Amrhein (1989) reviewed the literature on word reading versus picture naming and concluded that differences in practice cannot account for the word-reading advantage. Rather, they argued that the word-reading advantage relies on the fact that words can be read on the basis of their orthography alone, without access to the meaning code. In contrast, picture naming is usually assumed to require access to meaning.One approach to studying the processes underlying picture naming has been to vary individual item characteristics and observe which characteristics affect picture naming. Wingfield (1964, 1965) selected 26 pictures that varied widely in Thorndike-Lorge (1944) name frequency and found that naming latency was negatively correlated (r = -.80) with log frequency. Goodglass, Theurkauf, and Wingfield (1984) replicated this finding. White (1973a, 1973b) argued that age of acquisition might be a more predictive variable than word This research was supported by Research Grants AFOSR 89-0442 and F49620-92-J-0 119 from the Air Force Office of Scientific Research to 1.G.S.