According to Jones & Nisbett (1971). actors tend to explain their own behaviour in terms of situational factors, or characteristics of the object of their behaviour, whereas observers of the same behaviour tend to attribute it to characteristics of the actors themselves. Recently, Eiser, Sutton & Wober (1977) suggested that there may be a similar tendency in smokers' and non-smokers' views about cigarette smoking: non-smokers ('observers') may seek to explain smoking in terms of assumed characteristics of the average smoker, and smokers ('actors') may attach greater weight to factors such as the pleasure which smoking provides. The present report extends this approach, including additional comparisons between smokers' and non-smokers' attributions. programme appreciation diaries distributed by the Independent Broadcasting Authority, London. In all 1393 diaries were distributed to evaluate programmes in the week of 17 May 1976 to a representative sample of the population of N.W. England, aged 18 years and over. Of these 576 were drawn at random from the electoral register, the remainder having responded to previous surveys (including our own). A total of 367 usable diaries were returned (a fairly typical response rate), of which 314 were accompanied by usable questionnaires. As in our previous study, subjects were classfied by sex ( n for females, 170; for males, 144), age (18-34 years, n = 102; 35-54 years, n = 110; 55 years and over, n = 102) and social class (ABC,, n = 109; Clr n = 109; DE, n = 96), and also into smokers (n = 115), never-smokers (n = 117), and ex-smokers (n = 82) on the basis of the answers to a question asking if they had smoked cigarettes regularly during 1975, or had ever done so in the past. These figures imply some overrepresentation of ex-smokers and underrepresentation of smokers compared with proportions in the general population.All subjects were asked to rate, in terms of three categories ( I = no influence; 2 = some influence; 3 =great influence), how much influence they thought each of the following four factors might have in 'starting people off smoking': (a) having smokers in the family; (b) having close friends who smoke; (c) seeing advertisements about smoking on TV or anywhere else; and (d) seeing actors and personalities smoking on TV. Four questions were then asked in a slightly different form for smokers and non-smokers. Smokers were asked (in addition to a few other questions unrelated to this report): (e) 'How difficult would it be for you to give up cigarettes if you wanted to?' (Response categories: 1 = Not at all difficult, 2 = Fairly difficult; 3 = Very difficult); u) 'Would you like to stop smoking cigarettes altogether if you could do so easily?'