Background There is growing evidence that endothelial failure and subsequent blood brain barrier (BBB) breakdown initiate cerebral small vessel disease (CSVD) pathology. In spontaneously hypertensive stroke-prone rats (SHRSP) endothelial damage is indicated by intraluminal accumulations of erythrocytes (erythrocyte thrombi) that are not observed with current magnetic resonance imaging techniques. Two-photon microscopy (2 PM) offers the potential for real-time direct detection of the small vasculature. Thus, within this pilot study we investigated the sensitivity of 2 PM to detect erythrocyte thrombi expressing initiating CSVD phenomena in vivo. Methods Eight SHRSP and 13 Wistar controls were used for in vivo imaging and subsequent histology with haematoxylin-eosin (HE). For 2 PM, cerebral blood vessels were labeled by fluorescent Dextran (70 kDa) applied intraorbitally. The correlation between vascular erythrocyte thrombi observed by 2 PM and HE-staining was assessed. Artificial surgical damage and parenchymal Dextran distribution were analyzed postmortem. Results Dextran was distributed within the small vessel walls and co-localized with IgG. Artificial surgical damage was comparable between SHRSP and Wistar controls and mainly affected the small vasculature. In fewer than 20% of animals there was correlation between erythrocyte thrombi as observed with 2 PM and histologically with HE. Conclusions Contrary to our initial expectations, there was little agreement between intravital 2 PM imaging and histology for the detection of erythrocyte thrombi. Two-photon microscopy is a valuable technique that complements but does not replace the value of conventional histology.