The surface of the martian moon Phobos is characterized by parallel and intersecting grooves that bear resemblance to secondary crater chains observed on planetary surfaces. Murray (2011) has hypothesized that the main groove-forming process on Phobos is the intersection of Phobos with ejecta from primary impact events on Mars to produce chains of secondary craters. The hypothesis infers a pattern of parallel jets of ejecta, either fluidized or solidified, that break into equally-spaced fragments and disperse uniformly along-trajectory during the flight from Mars to Phobos. At the moment of impact with Phobos the dispersed fragments emplace secondary craters that are aligned along strike corresponding to the flight pattern of ejecta along trajectory. The aspects of the characteristics of grooves on Phobos cited by this hypothesis that might be explained by secondary ejecta include: their observed linearity, parallelism, planar alignment, pitted nature, change in character along strike, and a “zone of avoidance” where ejecta from Mars is predicted not to impact (Murray, 2011).To test the hypothesis we plot precise Keplerian orbits for ejecta from Mars (elliptical and hyperbolic with periapsis located below the surface of Mars). From these trajectories we: (1) set the fragment dispersion limits of ejecta patterns required to emplace the more typically well-organized parallel grooves observed in returned images from Phobos; (2) plot ranges of the ejecta flight durations from Mars to Phobos and map regions of exposure; (3) utilize the same exposure map to observe trajectory-defined ejecta exposure shadows; (4) observe hemispheric exposure in response to shorter and longer durations of ejecta flight; (5) assess the viability of ejecta emplacing the large family of grooves covering most of the northern hemisphere of Phobos; and (6) plot the arrival of parallel lines of ejecta emplacing chains of craters at oblique incident angles. We also assess the bulk volume of ejecta from martian impact events and the number of events that are necessary to supply sufficient bulk densities of secondary impactor fragments.On the basis of our analysis, we find that six major predictions of the Murray hypothesis are not consistent with a wide range of Mars ejecta emplacement models and observations as follows: (1) To emplace families of parallel grooves as observed in the most common features (grooves that manifest virtually no positional defects), and to reach the maximum geographic extent of Phobos, grid patterns of ejecta fragments must be produced with nearly identical diameters (often tens of thousands in number) and must launch with virtually zero rates of dispersion (<1mm/s and <1.0μrad in all degrees of freedom) into fixed patterns of arrays where fragment dispersion is referenced to a common datum point for the duration of flights from Mars to Phobos of up to 3h. (2) Half of the areal region observed as a “zone of avoidance” (where grooves are absent on the trailing orbital apex of Phobos) is directly exposed to ejecta trajectories from the surface of Mars, which suggests that the “zone of avoidance” is unrelated to ejecta trajectories. (3) Several families of grooves display groove segments that are observed in a region of Phobos that is shadowed from martian ejecta trajectories for flight durations up to 3h. Where the Murray hypothesis predicts the emplacement of groove families from a single ejecta plume, this strongly suggests that these families of grooves could not have been produced by martian impact ejecta. (4) To reach increasingly westerly locations of Phobos ejecta must travel in space for substantially longer flight times (up to 20X) which would produce substantially lower secondary crater densities on the anti-Mars side of Phobos and observably reduce their pit organization. This is not observed. (5) The largest family of grooves cannot be emplaced by any valid trajectory from Mars in its present day or ancient orbit. (6) If emplaced by grid patterns of ejecta, the irregular topography and small-body radius of Phobos would clearly disrupt groove family linearity and parallelism due to the preponderance of oblique incident angle impacts. However, when viewed from any position, the vast majority of groove families and individual grooves appear to completely avoid the effects of Phobos' morphology. Based on our analysis we conclude that the Murray hypothesis, that many Phobos grooves are formed by intersection of ejecta from craters on Mars, is not valid.