The aim of this dissertation is twofold. On one hand, I describe the use of tense in the second language of the deaf – written Swedish. There is no previous description of this. On the other hand, I operationalize the transitivity model (Hopper & Thompson 1980) and test its scope. I test in part whether transitivity as it is described by Hopper & Thompson (1980) can be used to explain the use of tense by the deaf in my study. The reason why I connect transitivity with the use of tense is that I would like to see whether a transitivity analysis of a text can replace classification into foreground and background. If so, a transitivity analysis should be a sufficient method for testing the discourse hypothesis, which maintains that second language learners use their emerging verb morphology to signal foreground and background rather than tense. The material I use for both these aims consists of a total of 129 essays written by deaf high school students. The results show that the percentage of clauses with an appropriate choice of tense in all three subsets of material is high. One interesting trend that could be discerned with regard to use of tense is that when informants use a tense that is appropriate for the genre, their command of that tense is better than when it is used in a genre for which it is less appropriate. Calculations testing the hypothesis of co-variance between the components in the transitivity model show in general that such co-variance exists. Nonetheless, the degree of this varies and is far from total. Taken together, the calculations made in this investigation provide support for the claim that these nine components co-vary with one another. The results also show that there is no correlation between the components and the use of tense. Judging from the results, the value of the different components (positive or negative) appears to have no importance in determining whether the choice of tense is appropriate or not. Nor is there any significant difference in the subsets of material for my investigation of whether the number of positive values that a clause has is relevant in classifying the choice of tense as appropriate.