Grasspea is one of the oldest cultivated species and was known as far back as 8000 years before Christ. The history of grasspea cultivation in Poland is not known. According to literature, grasspea first appeared in Poland (in the region of Podlasie) accompanying lentil as a weed as far back as 17th century, together with Tatar settlements. One of the conditions for a broader introduction of the species in Polish agriculture is genetic improvement of a number of undesirable characters. Apart from the unquestionable advantages, grasspea is also characterized by a number of less favourable characters which will be improved. Apart from effects of recombination, such a possibility is offered by mutation induction through the introduction of additional variability of features, represented by mutants. Mutagenesis in particular can be a valuable supplement to the conventional plant breeding method. It can be used to create additional genetic variability that may be utilized by the plant breeder in the development of new cultivars for specific purposes or with specific adaptabilities. The paper presents a multivariate approach to the estimation of genetic variability for quantitative traits of grasspea mutants (Lathyrus sativus L.). Material for the study involved 20 mutants and two cultivars: Derek and Krab. The mutants were obtained with the use of helium-neon laser light in combination with two chemomutagens: MNU (N-nitroso-N-methylurea) and sodium azide (NaN3). Plant height, height of the lowest pod, number of branches per plant, number of pods per plant, pod length, number of seeds per plant, weight of seeds per plant and 1000-seed weight were observed and analysed. The multivariate two-factor analysis of variance and related multivariate methods were applied. Canonical variate analysis was a useful statistical tool for clear identification of multivariate genetic variation of grasspea mutants.