According to the Complete Peerage and the current Oxford Dictionary of National Biography, Henry Stafford (we'll call him Harry from now on) was born on September 4, 1455, making him about thirty-five months younger than Richard, Duke of Gloucester, born on October 2, 1452. Henry's birthdate in September 1455 is consistent with this 1475 Act of Parliament, which refers to an annuity of 40 pounds granted by Henry VI to Henry's grandfather, the first Duke of Buckingham:
The king our sovereign lord, for certain considerations influencing him and because his cousin Henry, the present duke of Buckingham, cousin and male heir of the said late duke, that is to say, son of Humphrey, son of the said late duke, was very young at the time of the death of the said late duke [at the battle of Northampton on July 10, 1460], that is to say, only four years of age, and was lawfully taken into the wardship of our said sovereign lord . . .
A number of sources, however, age Harry by a year by giving him a birth date of 1454. This confusion (which runs rampant on the Internet) appears to have arisen originally because of another mistake: the report by some chroniclers that Harry's father, Humphrey Stafford, the Earl of Stafford, was killed or mortally wounded at the first battle of St. Albans in May 1455. Since Harry is known to have had a younger brother, who was made a Knight of the Bath along with Harry and who was raised in Elizabeth Woodville's household with him, historians, like C.A.J. Armstrong in a note to his 1936 edition of Mancini, moved Harry's birth date back a year in order to avoid the impossible situation of the Earl of Stafford posthumously begetting Harry's younger brother.
But the Earl of Stafford did not in fact die at St. Albans, though he was wounded there. Armstrong himself had realized this by 1960, when he published his seminal article "Politics and the Battle of St. Albans, 1455." Instead, as Armstrong noted, the earl died in late 1458 of plague, according to a manuscript, Rawlinson B 355, cited by Armstrong in note 5 to his article. (Footnotes are a researcher's best friend.) Indeed, although some writers have depicted the Earl of Stafford as having been permanently prostrated by his wounds at St. Albans, he continued to take an active part in affairs following the 1455 battle: he was appointed to a commission of oyer and terminer in 1456 and to Edward of Lancaster's council and to other commissions in 1457. At some point in between sitting on commissions, the earl fathered Harry's younger brother.
So with the Earl of Stafford having survived St. Albans and lived long enough to father both Harry and his younger brother, the difficulty with Harry's birth date of 1455 disappears. Might we someday hear the sweet music of all of those erroneous online statements of "1454" being corrected? Probably not, but at least I've done my part.