Chitosan grafting onto silk was tested with three crosslinking agents: trifunctional epoxy resin Araldite DY-T, PEG400 dimethacrylate, and glutaraldehyde in acetic as well as in tartaric acid solutions. Operating conditions were studied to obtain a significant silk weighting with satisfactory graft yields. With the epoxy crosslinker the weight gain was in the range from 1.8 to 8.8 % with graft yield from 8 to 23 %. In the case of PEG400DMA weight gain was 8-12 % with 22-24 % graft yield. With glutaraldehyde in tartaric acid solution a maximum weight gain of 8.4 % with 27.6 % graft yield was obtained. Results of determination of primary amino groups on the grafted silk showed that with epoxy and glutaraldehyde, unlike PEG400DMA, the amino groups of chitosan were only partially involved in crosslinking. Results of DSC analyses suggested that the modification of fibroin structure in chitosan-grafted silk was stronger with glutaraldehyde than with epoxy or dimethacrylate. FTIR-ATR spectra of grafted fibers were found very similar to that of control silk with an additional weak peak ascribable to chitosan in 1180-1080 $cm^{-1}$ range. Surface investigation through AFM showed clear morphology differences between chitosan-grafted silk with epoxy or dimethacrylate and that crosslinked with glutaraldehyde; the latter appears uneven and scale-like, the others slightly rougher than the original silk.