Purpose. To investigate variability of choriocapillaris blood flow pat
terns. Methods. After the intravenous injection of indocyanine green,
angiograms were recorded at 30 images per second in rhesus monkey eyes
using a fundus camera equipped with a pulsed laser diode light source
, synchronized with a gated (5 msec), intensified charge-coupled devic
e, or CCD, video camera. Images of choriocapillaris filling alone were
extracted. Plastic corrosion casts were made of two of the monkey's c
horoidal vasculatures for subsequent scanning electron microscopy exam
ination. Results. Pulsed laser indocyanine green fluorescence excitati
on produced better definition of choriocapillaris filling than had bee
n achieved using continuous illumination. No correlation was found bet
ween the choriocapillaris plexus architecture revealed by the plastic
corrosion casts and the observed choriocapillaris lobular filling. Ove
rall posterior pole choriocapillaris dye-filling patterns were relativ
ely stable for periods of days, but they changed gradually for periods
of weeks. Localized minor pattern changes occurred on a much shorter
time scale. Choriocapillaris filling patterns were altered by acutely
elevating intraocular pressure, by O2 and CO2 breathing, and by argon
laser retinal photocoagulation of adjacent areas. Conclusions. Chorioc
apillaris filling patterns appear to be determined by the network of p
erfusion pressure gradients that exist among the interspersed feeding
arterioles and draining venules connected to the choriocapillaris plex
us. Changes in intraocular pressure and in blood P-O2 and P-CO2 levels
can produce marked changes in the distribution of choriocapillaris bl
ood flow. Retinal laser photocoagulation of adjacent fundus areas alte
rs choriocapillaris blood flow to the extent that the altered flow mig
ht be an important factor in the beneficial results attributed to reti
nal laser treatment.